Links between Signaling, Transcriptional Cascades, Epithelial Remodeling, and Proliferation in the Hair Bud
Previously, we discovered that early during hair follicle morphogenesis, E-cadherin gene expression is down-regulated concomitantly with the invagination of developing bud cells into the skin [4]. Because the timing of this event correlated with the activation of a LEF-1/β-catenin transcription factor complex [20], we were intrigued by the presence of a putative LEF-1/TCF binding site in the E-cadherin promoter. This prompted an investigation that subsequently led to our discovery that LEF-1/β-catenin can contribute to repression of E-cadherin gene expression in skin keratinocytes [4]. In the course of these studies, we also noted that Snail can also contribute to this process in keratinocytes in vitro, and our present studies revealed that Snail is expressed at the right place and time to be physiologically relevant in the process.
In noggin-null embryonic skin, LEF-1 expression and subsequent activation of the LEF-1/β-catenin reporter gene is abrogated in the developing placodes. The corresponding failure of E-cadherin down-regulation underscores the importance of Wnt/noggin signaling in regulating this event in follicle morphogenesis [4]. Conditional gene targeting studies will be necessary to establish whether Snail family members also contribute to the down-regulation in E-cadherin gene expression that occurs during follicle formation. However, it is intriguing that K14-Snail Tg epidermis displayed a marked down-regulation in E-cadherin expression, thereby demonstrating its potential to do so in skin. Our prior findings showed that by elevating E-cadherin levels or by conditionally ablating α-catenin, hair follicle morphogenesis can be impaired [4,7]. The marked epidermal hyperproliferation seen in the K14-Snail Tg skin, coupled with the converse suppression of proliferation and Snail in TGF-β2-null hair buds led us to wonder whether the down-regulation of E-cadherin during follicle morphogenesis might have a direct impact on elevating the proliferative state of these cells.
Our Tg studies suggested that, at least in part through its regulation of E-cadherin, Snail is able to influence the subcellular localization of a variety of AJ-associated proteins. One of these appears to be Ajuba, which was previously shown to have the dual capacity to bind Grb-2 as well as α-catenin [9,10]. Our studies revealed that in skin keratinocytes that either harbor a conditional null mutation in α-catenin or that overexpress Snail, Ajuba develops an interaction with Grb-2 that is otherwise not observed in WT keratinocytes. The corresponding abilities of either Snail-transfected or Ajuba-transfected keratinocytes to exhibit elevated activation of the Ras-MAPK pathway suggest that the Grb-2 association of Ajuba under conditions of reduced levels of AJ proteins may be directly relevant to the parallel in hyperproliferation.
In stable epithelial (i.e., Madin-Darby canine kidney, or MDCK) cell lines, Snail has been shown to block cell cycle progression and promote motility and shape changes for invasion [47]. While our in vivo studies are consistent with a role for Snail in motility and epithelial remodeling, they differ with respect to Snail's apparent proliferative effects. A priori, this could be simply due to variations in the response of different cell types to Snail expression. Alternatively, these differences may be relevant to the benefit of using mouse models to reveal functions not always recapitulated in stable cell line models. Future studies should highlight the underlying reasons for these opposing results.
Irrespective of these differences, our in vivo studies do not stand alone, as there are many situations in which a down-regulation in AJ proteins correlate with enhanced proliferation. In fact, a myriad of diverse mechanisms have been implicated in activating epithelial proliferation upon down-regulation of AJ proteins [7,23,24,48]. Sifting through these converging pathways is likely to be a difficult and painstaking process. This said, by identifying the status of different players involved in specific cell types and at specific stages in development, our mechanistic understanding of how intercellular remodeling is linked to proliferation in epithelial morphogenesis should begin to surface in the future. Elucidating the molecular mechanisms through which these networks converge is also a prerequisite for understanding how these processes go awry during tumorigenesis.